Trait space::KnnMap

pub trait KnnMap: KnnPoints {
    type Value;
    fn get_value(&self, index: Self::Ix) -> &Self::Value;

    fn knn_values(
        &self, 
        query: &Self::Point, 
        num: usize
    ) -> Vec<(Neighbor<<Self::Metric as Metric<Self::Point>>::Unit, Self::Ix>, &Self::Value)> { ... }
    fn nn_value(
        &self, 
        query: &Self::Point
    ) -> Option<(Neighbor<<Self::Metric as Metric<Self::Point>>::Unit, Self::Ix>, &Self::Value)> { ... }
    fn knn_keys_values(
        &self, 
        query: &Self::Point, 
        num: usize
    ) -> Vec<(Neighbor<<Self::Metric as Metric<Self::Point>>::Unit, Self::Ix>, &Self::Point, &Self::Value)> { ... }
    fn nn_key_value(
        &self, 
        query: &Self::Point
    ) -> Option<(Neighbor<<Self::Metric as Metric<Self::Point>>::Unit, Self::Ix>, &Self::Point, &Self::Value)> { ... }
}

This trait gives knn search collections the ability to give the nearest neighbor values back.

This is not the final API. Eventually, the iterator type will be chosen by the collection, but for now it is a Vec until Rust stabilizes GATs.

Associated Types

type Value

Required methods

fn get_value(&self, index: Self::Ix) -> &Self::Value

Get a value using a neighbor index returned by Knn::knn or Knn::nn.

This should only be used directly after one of the mentioned methods are called to retrieve a value associated with a neighbor, and will panic if the index is incorrect due to mutating the data structure thereafter. The index is only valid up until the next mutation.

Provided methods

fn knn_values(
    &self,
    query: &Self::Point,
    num: usize
) -> Vec<(Neighbor<<Self::Metric as Metric<Self::Point>>::Unit, Self::Ix>, &Self::Value)>

Get num nearest neighbor keys of target.

For many KNN search algorithms, the returned neighbors are approximate, and may not be the actual nearest neighbors.

fn nn_value(
    &self,
    query: &Self::Point
) -> Option<(Neighbor<<Self::Metric as Metric<Self::Point>>::Unit, Self::Ix>, &Self::Value)>

Get the nearest neighbor key of target.

For many KNN search algorithms, the returned neighbors are approximate, and may not be the actual nearest neighbors.

fn knn_keys_values(
    &self,
    query: &Self::Point,
    num: usize
) -> Vec<(Neighbor<<Self::Metric as Metric<Self::Point>>::Unit, Self::Ix>, &Self::Point, &Self::Value)>

Get num nearest neighbor keys of target.

For many KNN search algorithms, the returned neighbors are approximate, and may not be the actual nearest neighbors.

fn nn_key_value(
    &self,
    query: &Self::Point
) -> Option<(Neighbor<<Self::Metric as Metric<Self::Point>>::Unit, Self::Ix>, &Self::Point, &Self::Value)>

Get the nearest neighbor key of target.

For many KNN search algorithms, the returned neighbors are approximate, and may not be the actual nearest neighbors.

Implementors